The aims of this research are to: 1) assess the long-term effectiveness of a successful smoking prevention program; 2) explore the differential pattern of program effects over time; 3) complete more in-depth analyses of whom the program affected and whom it did not affect, especially over time; 4) assess the effects of repeated testing on students not exposed to a smoking prevention program; and 5) assess [program effects on the reported level of smoking by parents, siblings, and friends of students exposed to the program. These aims will be met by the collection of five- and six-year follow-up data from students who participated in the original Waterloo Smoking Prevention Study and a comparable sample of same-age students not included in the original study. The Waterloo Study worth long-term follow-up because it represents the strongest test to date of the social influences approach to smoking prevention. A large enough number of schools were assigned (mostly on a random basis) to ensure (a) pretest comparability, and (b) that the classroom (the unit of intervention) could be used as the unit of analysis (at least for major tests of program effectiveness). Results up to 2 1/2 years after the core intervention demonstrated that the Waterloo program was at least as effective as others that have been reported in the literature. This follow-up study will be conducted at two sites: data will be collected and some preliminary and nonparametric analyses will be conducted at the University of Waterloo by the Co-P.I. and Co-Investigators (on a subcontract-consortium basis), but the bulk of the statistical analyses and report generation will be conducted at the University of Southern California by the P.I. and staff of the Health Behavior Research Institute.